Short- and Long-Term Patient Outcomes From Combined Coronary Endarterectomy and Coronary Artery Bypass Grafting: A Meta-Analysis of 63,730 Patients (PRISMA).

This meta-analysis aimed to compare the short- and long-term outcomes in patients undergoing combined coronary endarterectomy and coronary artery bypass grafting (CE + CABG) versus isolated CABG, and particularly to examine subgroup patients with high-risk profile and patients with diffuse disease in the left anterior descending artery (LAD).Studies published between January 1, 1970 and May 31, 2015 were searched in the literature databases, including Ovid Medline, Embase, PubMed, and ISI Web of Science.A total of 30 eligible studies including 63,730 patients were analyzed.Five authors extracted data from the included studies independently.Meta-analysis on the total patients revealed that CE + CABG was associated with significantly increased 30-day postoperative all-cause mortality compared with isolated CABG (OR = 1.86, 95% CI: 1.66-2.08, z = 10.99, P < 0.0001). Subgroup analysis on patients with high-risk profile and patients with diffuse disease in the LAD showed that 30-day mortality after CE + CABG was 2.6 folds (OR = 2.60, 95% CI: 1.39-4.86, z = 2.99, P = 0.003) and 3.93 folds (OR = 3.93, 95% CI: 1.40-11.0, z = 2.60, P = 0.009) of that after isolated CABG in the respective subgroup. In contrast, the mortality was comparable in CE + off-pump CABG and CE + on-pump CABG groups (OR = 0.53, 95% CI: 0.18-1.55, z = 1.16, P = 0.248). In addition, the incidences of perioperative myocardial infarction (MI) and 30-day postoperative complications, including low output syndrome (LOS), MI, ventricular tachycardia (VT), and renal dysfunction after CE + CABG were significantly higher than those after isolated CABG (all P < 0.05). In high-risk patient subgroup, CE + CABG significantly increased the incidences of postoperative LOS, MI, and renal function compared with isolated CABG (all P < 0.05). The incidence of perioperative myocardial after CE + CABG was 2.86 and 2.92 times of that after isolated CABG in high-risk patients and patients with diffuse disease in LAD, respectively. Analysis on the recent reports (published later than 2000) showed consistent results as the analysis including all the eligible reports. Long-term survival was comparable in CE + CABG and isolated CABG groups (hazardous ratio = 1.16, 95% CI: 0.32-4.22, z = 0.23, P = 0.819).CE + CABG appears to be associated with poor short-term outcomes, particularly in high-risk patients and patients with diffuse disease in the LAD.

INTRODUCTION

Coronary artery bypass grafting (CABG) is 1 of the most common surgical interventions for coronary revascularization.[1] However, it often fails to achieve satisfactory outcomes in patients with a high-risk profile, which is characterized by old age, severe left ventricular dysfunction, diffuse coronary artery disease (ie, at least 75% of the segment distal to the lesion has a vessel diameter of <2 mm),[2] comorbidities such as diabetes and peripheral vascular disease, and/or previous percutaneous coronary intervention (PCI).[3,4] Coronary endarterectomy (CE) combined with CABG (CE + CABG) had been commonly practiced in the past.[5] Theoretically, simultaneous removal of plaques in the target vessels to CABG can achieve complete revascularization, particularly for patients with diffuse coronary artery disease, who usually have limited available target vessels for CABG. However, in recent years, surgeons become increasingly cautious toward CE + CABG.[6] It has been reported that patients undergoing CE + CABG showed poorer 30-day postoperative outcomes than patients undergoing isolated CABG.[7,8] Soylu et al conducted a meta-analysis and found that CE + CABG increased early postoperative mortality and the incidence of perioperative myocardial infarction (MI).[5]

Comparison of 30-day postoperative outcomes from CE + CABG versus isolated CABG has been conducted on patients with unselected risk of coronary artery disease (CAD).[5] Patients undergoing CE usually have high-risk profile, such as having diffuse coronary artery disease and poorer physical condition, compared with patients undergoing isolated CABG.[1] The patient preoperative clinical characteristics may not match in CE + CABG and isolated CABG groups in previous meta-analysis.[5] Thus, meta-analysis on patients with unselected risk might not accurately reveal the benefits or adverse effects of CE procedure because a poor outcomes from CE + CABG may be associated with poor patient preoperative clinical characteristics. Furthermore, the effect of CE + CABG on long-term survival is still unclear. This study aimed to compare the short- and long-term outcomes from CE + CABG versus isolated CABG, particularly in patients with high-risk profile and patients with diffuse disease in the left anterior descending artery (LAD).

METHODS

Guidelines and Data Collection

As this meta-analysis, as a systematic review, does not involve animal experiments or direct human trials, there is no need to conduct special ethic review and the ethical approval is not necessary. The design of this meta-analysis stringently followed the guidelines for Quality of Reporting of Meta-analysis.[8,9] We systematically searched Ovid Medline, Embase, PubMed, and ISI Web of Science for relevant studies published between January 1, 1970 and May 31, 2015. The search terms were “CABG,” “endarterectomy,” “left anterior descending,” “high risk,” “off-pump CABG,” and “diffuse coronary artery disease.” The reference lists of selected articles, conference proceedings, and personal files for relevant citations were also screened. Publication language was not restricted in the search.

Study Inclusion and Exclusion Criteria

Eligible studies met the following inclusion criteria: patients underwent CABG; CE + CABG group and isolated CABG group were compared; CE + off-pump CABG group and CE + on-pump CABG group were compared; short-term outcomes included 30-day postoperative mortality, low output syndrome (LOS), perioperative MI, postoperative renal dysfunction, postoperative MI, postoperative arrhythmia including atrial fibrillation (AF), and ventricular tachycardia (VT) were studies; long-term patient survival was investigated. The exclusion criteria were duplicate of previous publication; abstract, review, comment, and editorial; preclinical studies; missing original data; lack of isolated CABG as a control group; patients undergoing other procedures such as valve replacement and major vascular surgery in addition to CE and CABG.

Diffuse Coronary Artery Disease and High-Risk Definitions

According to the 2014 ESC/EACTS guidelines on myocardial revascularization, the definition of diffuse coronary artery disease in this study was that at least 75% of the segment distal to the lesion has a vessel diameter <2 mm.[2] The definition of high-risk patients in both CE + CABG and isolated CABG groups was that patients met 1 or more of the following criteria: ages ≥70 years; diffuse disease in at least 1 coronary artery; left main stem stenosis ≥70%; preoperative comorbidities such as diabetes, peripheral vascular disease, LOS, or MI; previous PCI or CABG; unstable angina (CCS class III and IV) or poor left ventricular function (LVEF ≤40%) as assessed by left ventriculography.

Data Extraction and Quality Assessment

Five reviewers (JW, WY, CG, MG, and YY) extracted the following data: details of the publication (first author's surname, trial acronym, enrollment period, and year of publication), patient inclusion and exclusion criteria, patient demographics, sample size in control and study groups, analyses on high-risk patients, patients with diffuse diseased in the LAD, comparison of CE + off pump CABG versus CE + on-pump CABG, and outcomes including 30-day postoperative mortality, LOS, AF, VT, and renal dysfunction, perioperative and postoperative MI, and long-term survival. When the 5 reviewers had different opinions toward some studies, the problematic studies were reviewed again and the disagreement was discussed openly until a consent was reached. Downs and Black score was used to evaluate the quality of each included study.[9] The checklist of Downs and Black score consists of 26 items distributed in 5 sub-scales: reporting (9 items), which estimates whether the information provided in the paper is sufficient to allow a reader to make an unbiased assessment of the findings of the study; external validity (3 items), which addresses the extent to which the findings from the study could be generalized to the population from which the study subjects were derived; bias (7 items), which addresses biases in the measurement of the intervention and the outcome; confounding (6 items), which addresses bias in the selection of study subjects; power (1 item), which attempts to assess whether the negative findings from a study could be due to chance. Except for 1 item in the reporting subscale, which was scored 0 to 2, and the single item in the power subscale, which was scored 0 to 5, all the answers to other items were scored 0 or 1. The total maximum score was therefore 31. Each reviewer scored the included studies separately, and an average score was calculated and used as the final Downs and Black score for the study. A score ≥20 represents satisfactory quality, and a score <20 represents poor quality. When the 5 reviewers have different opinions toward some studies, the problematic studies were reviewed again and the disagreement was discussed openly until a consent was reached.

Primary and Secondary Endpoints

The primary endpoint of this study was the odds ratio (OR) of 30-day postoperative all-cause mortality after CE + CABG versus isolated CABG (including subgroup analysis of 30-day mortality in high-risk patients and patients with diffuse disease in the LAD) and OR of 30-day mortality after CE + off-pump CABG versus CE + on-pump CABG. The secondary endpoints were hazard ratio (HR) of long-term survival after CE + CABG versus isolated CABG and the OR of the incidences of 30-day postoperative LOS, MI, AF, VT, renal dysfunction, and perioperative MI after CE + CABG versus isolated CABG. LOS was defined as a requirement of intra-aortic balloon pumping or infusion of norepinephrine, epinephrine, milrinone, or dobutamine after the first hour postoperatively to maintain systolic blood pressure ≥90 mm Hg. Postoperative or perioperative (as classified by the authors) MI was defined as creatine kinase–MB increase >80 μg/L or troponin T >3.0 μg/L the first 48 hr after surgery. The definition of 30-day postoperative MI was enzymatic elevation of creatine kinase–MB >10 μg/L or troponin T >0.1 μg/L together with at least 1 of the following findings: classic angina symptoms, electrocardiograph (ECG) signs of necrosis or ischemia, or coronary re-intervention. Postoperative AF was defined by the documentation of AF of any duration at any time during the postoperative period on a physician assessment, on the basis of a rhythm strip or 12-lead electrocardiogram recording. Postoperative VT was defined as the uniform tachyarrhythmia of ventricular origin that had a duration of >20 sec after surgery, unless terminated earlier because of hemodynamic collapse. Renal dysfunction was defined as a need of acute hemodialysis, blood creatinine level ≥200 μg/L, or blood creatinine 2 times the preoperative value.

Data Analysis

The statistical analysis software Stata (version 13.1) was used for data analysis. P value was 2-sided and P < 0.05 was considered statistically significant. Study heterogeneity was evaluated by I2 index. I2 = 25% to 49%, 50% to 74%, and ≥75% was considered low, moderate, and high heterogeneity, respectively. I2 ≥ 50% was considered significantly heterogeneous. A fixed-effect model was used for meta-analysis without significant study heterogeneity. A random-effect model was used when significant heterogeneity was found in the included studies. ORs with 95% CI were reported. HR was used to assess data on time-to-event and survival. Parmar et al[10] introduced the standard approaches to extract data on time-to-event analyses. Publication bias was examined by Funnel plot and Egger's weighted regression. When the P value in the Egger's test was <0.05, the publication bias was consider significant.

RESULTS

Characteristics and Quality of the Included Studies

The flow chart for study screening is shown in Figure 1. A total of 2230 articles were retrieved by using the search terms; among them, 2192 records were excluded for the reasons listed in Figure 1. Full-text reports were obtained for 38 studies. Of the 38 reports, 5 lack control group; 3 contain overlapping data. Thus, a total of 30 observational studies, including 63,730 patients, were analyzed (Fig. 1 and Table 1    ).[1,6-8,11-36] Patient characteristics, endpoints in the included studies, sample size of both groups, and Downs and Black Score are presented in Table 1    . Of the 30 eligible reports, 27 compared 30-day mortality after CE + CABG versus isolated CABG; 5 compared 30-day mortality after CE + off-pump CABG versus CE + on-pump CABG; 10 investigated high-risk patients undergoing CABG with or without CE (the patient preoperative clinical characteristics were matched in CE + CABG and isolated CABG groups); 5 focused on patients with diffuse disease in the LAD undergoing CABG with or without CE (Table 1    ). The Downs and Black Score of the 30 articles was between 20 and 24 (Table 1    ), suggesting that the quality of all the 30 articles was satisfactory. Long-term survival data in the included studies, such as follow-up time, follow-up ratio, and survival, are presented in Table 2.

FIGURE 1

Study selection flow chart.

TABLE 1

Characteristics of the Included Studies

TABLE 1 (Continued)

Characteristics of the Included Studies

Primary Endpoints

Meta-analysis on patients with unselected risk of CAD demonstrated that CE + CABG was associated with significantly increased 30-day postoperative all-cause mortality compared with isolated CABG (OR = 1.86, 95% CI: 1.66–2.08, z = 10.99, P < 0.0001, Fig. 2) without study heterogeneity (I2 = 0.0%, P = 0.555). Meta-analysis including only the recent reports (published later than 2000) showed consistent results (OR = 1.82, 95% CI: 1.47–2.25, z = 5.57, P < 0.0001, Fig. 2) as the meta-analysis including all the eligible studies. Subgroup analysis on high-risk patients revealed similar results with even higher extent of increase in the mortality in CE + CABG group compared with isolated CABG group (OR = 2.60, 95% CI: 1.39–4.86, z = 2.99, P = 0.003, Fig. 3A) without study heterogeneity (I2 = 0.0%, P = 0.924), and consistent results were obtained from the meta-analysis including only the recent reports (OR = 3.10, 95% CI: 1.32–7.30, z = 2.59, P = 0.009, Fig. 3A). Subgroup analysis of patients with diffuse disease in the LAD showed that the mortality in LAD-CE + CABG group was 3.93 times of that in isolated CABG group (OR = 3.93, 95% CI: 1.40–11.00, z = 2.60, P = 0.009, Fig. 3B), and analysis on the recent reports revealed consistent results (OR = 3.66, 95% CI: 1.02–13.19, z = 1.99, P = 0.047, Fig. 3B). Sensitivity analysis found no significant impact of each individual study on the overall results. Neither funnel plot (Fig. 4) nor Egger's weighted regression showed significant publication bias (P = 0.546). Furthermore, patients undergoing CE + off-pump CABG and patients undergoing CE + on-pump CABG showed similar 30-day postoperative mortality (OR = 0.53, 95% CI: 0.18–1.55, z = 1.16, P = 0.248, Fig. 5).

FIGURE 2

Analysis of 30-day mortality after CE + CABG versus isolated CABG. CABG = coronary artery bypass grafting, CE = coronary endarterectomy, CI = confidence interval, OR = odds ratio.

FIGURE 3

Subgroup analysis of 30-day postoperative mortality. (A) Subgroup analysis of 30-day mortality in high-risk patients. (B) Subgroup analysis of 30-day postoperative mortality in patients with diffuse disease in the LAD. CI = confidence interval, LAD = left anterior descending artery, OR = odds ratio.

FIGURE 4

Funnel plot for the assessment of publication bias on 30-day postoperative all-cause mortality.

FIGURE 5

Analysis of 30-day mortality after CE + off-pump CABG versus CE + on-pump CABG. CABG = coronary artery bypass grafting, CE = coronary endarterectomy, CI = confidence interval, OR = odds ratio.

Secondary Endpoints

In contrast to 30-day postoperative mortality, long-term survival was comparable in CE + CABG and isolated CABG groups (HR = 1.16, 95% CI: 0.32–4.22, z = 0.23, P = 0.819, Fig. 6). Analysis of the recent reports (published later than 2000) showed consistent results (HR = 1.47, 95% CI: 0.17–13.14, z = 0.35, P = 0.728, Fig. 6). The studies included for long-term survival analysis showed no study heterogeneity (I2 = 0.0%, P = 0.999) and no significant publication bias (P = 0.057).

FIGURE 6

Analysis of long-term survival after CE + CABG compared with isolated CABG. CABG = coronary artery bypass grafting, CE = coronary endarterectomy, CI = confidence interval, HR = hazard ratio.

Meta-analysis on the incidences of 30-day postoperative complications revealed that compared with patients undergoing isolated CABG, patients undergoing CE + CABG exhibited significantly higher incidences of perioperative MI and 30-day postoperative LOS, MI, VT, and renal dysfunction with various study heterogeneity (Table 3). The study heterogeneity was particularly high for postoperative MI (Table 3). The incidence of AF was similar in the 2 groups of patients (Table 3) without significant heterogeneity.

TABLE 1 (Continued)

Characteristics of the Included Studies

Subgroup analysis on high-risk patients revealed that the incidence of postoperative MI in CE + CABG group was 2.02 times of that in isolated CABG group (Table 3) without study heterogeneity. Thus, 1 of the sources of heterogeneity for postoperative MI in the analysis on all the patients was probably attributable to reports including patients with unselected risk of CAD. In addition, in the high-risk patient subgroup, the incidences of perioperative MI and 30-day postoperative renal dysfunction and LOS in CE + CABG group were significantly higher than those in isolated CABG group (Table 3). The incidence of 30-day postoperative VT was similar in the high-risk patients undergoing CE + CABG versus isolated CABG (Table 3). In the subgroup analysis of patients with diffuse disease in the LAD, the incidence of perioperative MI in LAD-CE +CABG group was 2.92 times of that in isolated CABG group (Table 3).

Meta-analysis on the recent studies (published later than 2000) showed that patients undergoing CE + CABG exhibited higher incidences of perioperative MI and 30-day postoperative LOS, MI, VT, and renal dysfunction compared with patients undergoing isolate CABG (Table 4). These results are consistent with the results from the analysis including all the eligible reports (Table 3). Subgroup analysis of high-risk patients demonstrated that the incidences of perioperative MI, 30-day postoperative LOS, and MI in CE + CABG group was 2.56, 2.32, and 1.95 folds of those in isolated CABG group, respectively, without significant heterogeneity (Table 4).

TABLE 1 (Continued)

Characteristics of the Included Studies

DISCUSSION

CE combined with CABG to achieve complete revascularization has been practiced for more than 40 years.[6,18] However, CE has been found to be associated with poor 30-day postoperative outcomes,[14,37] which seriously dampens the enthusiasm of cardiac surgeons toward this procedure.[6] Similar to previous reports,[5] this present study also showed that CE + CABG was associated with significantly increased 30-day postoperative all-cause mortality and higher incidences of postoperative LOS, MI, VT, and renal dysfunction. Meta-analysis only including the recent eligible reports (published later than 2000) showed consistent results as the meta-analysis including all the eligible studies. To rule out the possibility that the poor short-term outcomes after CE + CABG might be associated with the poor physical condition of patients in CE + CABG group,[8] in this study, we performed subgroup analysis on high-risk patients with matched preoperative clinical characteristics in CE + CABG and isolated CABG groups. We found that 30-day mortality and incidences of postoperative complications after CE + CABG were dramatically higher than those after isolated CABG in high-risk patients. Similar results were observed in the subgroup analysis of patients with diffuse disease in LAD. These results indicate that CE procedure may result in poor short-term postoperative outcomes, particularly in patients with high-risk profile or with diffuse disease in the LAD.

The poor short-term postoperative outcomes in patients undergoing CE + CABG, particularly in high-risk patients, might result from endothelium damage during CE. Intact coronary endothelium can produce vasoactive factors to neutralize leukocyte adhesion and platelet aggregation, consequently reducing inflammation and thrombosis in blood vessels.[38] The damage caused by CE leads to endothelium dysfunction. Acute perioperative MI, a key complication associated with CE failure, might be associated with thrombosis, which can be induced by removal of the endothelial lining during CE.[30] In addition, the time of cardiopulmonary bypass (CPB) during CE + CABG is usually longer than that during isolated on-pump CABG, consequently deteriorating CPB-induced ischemia-reperfusion injury (IRI) and leading to poor outcomes.[7] Endothelium damage and IRI reduce or inactivate vasoactive factors, consequently increasing neutrophil–endothelium adhesion, thrombosis, inflammation, and the incidences of postoperative complications.[39] Furthermore, MI caused by residual lesion-induced occlusion also contribute to the postoperative morbidity and mortality of CE.[40,41]

In this report, we also performed subgroup analysis on patients with diffuse disease in the LAD. The effects of LAD-CE + CABG on patient outcomes are still controversial. Byrne et al[42] showed that patients undergoing LAD-CE + CABG had low hospital mortality and morbidity. On the other hand, Silberman et al[7] suggested that patients with diffuse disease in the LAD might have a high risk for perioperative MI and death after CE. Our results also showed that LAD-CE + CABG dramatically increased 30-day postoperative mortality and incidence of 30-day postoperative complications compared with isolated CABG. The increased 30-day postoperative mortality associated with LAD-CE might result from residual lesions in the LAD. Lawrie et al[40] investigated the independent influence of site of residual disease on late survival and found that residual lesions in the LAD artery or circumflex coronary arteries were significant predictors of survival, whereas residual lesions in the right coronary artery exerted a lesser effect. Our study also found that short-term postoperative mortality was similar in CE + off-pump CABG and CE + on-pump CABG groups. Similarly, Qiu et al[34] also showed that on-pump and off-pump CE appeared to have similar early and mid-term outcomes. Sirivella et al[6] reported that on-pump versus off-pump CE technique did not alter patient outcomes.

Although short-term outcomes from CE + CABG appeared to be poorer than those from isolated CABG, our findings demonstrated that long-term survival was comparable in CE + CABG and isolated CABG groups. Previous reports also demonstrated that long-term survival was similar in CE + CABG and CABG groups.[1,30] Since patients requiring CE + CABG usually have high-risk diffuse CAD,[1] a comparable survival in CE + CABG and isolated CABG groups may indicate the beneficial effects of complete revascularization of CE + CABG. However, long-term complications, such as MI, angina, heart failure, and cardiovascular mortality, are still unknown for patients undergoing CE + CABG. Soylu et al[5] showed that the angiographic patency at follow-up was worse in CE + CABG group. Moreover, in this present study, the meta-analysis to compare long-term survival between CE + CABG group and isolated CABG group was on patients with unselected risks, and a subgroup analysis in patients with high-risk profile was not feasible because of lacking sufficient eligible reports. Thus, whether CE procedure can exert long-term survival benefit in high-risk patients remain inconclusive and require further investigations.

Surgeons should assess the benefits and adverse effects of CE adequately and optimize the surgical strategy to achieve satisfactory clinical outcomes. Guidelines for CE are currently not available. Fukui et al[1] recommended CE as the only revascularization method suitable for diffusely diseased vessels with either continuous calcified, soft, or hard fibrous plaques. It has been found that removal of the endothelial lining or residual obstruction caused postoperative morbidity and mortality after CE.[30,41] Our results also showed that in high-risk patients, the incidence of perioperative MI after CE + CABG was markedly higher than that after isolated CABG. Thus, we believe that unnecessary CE should be avoided. CE should be only used on patients who have diffuse CAD and whose CAD cannot be treated effectively by CABG alone. Because incomplete revascularization is 1 of the most critical factors influencing long-term mortality and morbidity,[40] to achieve complete revascularization, CE is recommended as an adjunct to CABG for this group of high-risk patients.[1,6] For patients who require CE, surgeons should maximally reduce the risk of postoperative complications, such as thrombosis, by using proper surgical techniques and aggressive postoperative anticoagulant and antiplatelet therapies.[5] Soylu et al[43] found that open-CE + CABG appeared to reduce 30-day mortality compared with closed-CE + CABG in patients with diffuse coronary artery disease. Tasdemir et al[22] suggested that compared with closed-CE, open-CE, in which the arterial lumen can be exposed, better facilitate a complete removal of plaques and avoid postoperative thrombotic occlusion. Nishi et al[44] reported that the incidences of postoperative MI, intra-aortic balloon pump use, and stroke were lower with open-CE than closed-CE. Thus, to achieve adequate revascularization in high-risk patients, surgeons should consider patient clinical presentation, extent and relevance of ischemia, and other comorbidities when optimizing a therapeutic strategy.[45]

To our best knowledge, this is the first meta-analysis focusing on high-risk patients to compare the short- and long-term outcomes from CE + CABG versus isolated CABG. In the subgroup analysis of high-risk patients, the preoperative clinical characteristics were matched in CE + CABG and isolated CABG groups. Thus, this current meta-analysis may accurately reveal the benefits (possible long-term survival benefit owing to a complete revascularization) or adverse effects (poor short-term outcomes) associated with CE procedure. Furthermore, this is also the first meta-analysis to confirm that CE + CABG is associated with dramatically higher 30-day postoperative mortality compared with isolated CABG in patients with diffuse disease in the LAD. The consistent results from the meta-analysis only including the recent eligible studies (published later than 2000) and from the meta-analysis including all the eligible studies suggest that evolution of surgical technology and patient care appear not affect the study endpoints significantly, further confirm the poor short-term outcomes associated with CE + CABG.

The major limitation of this meta-analysis is that all the included studies are observational studies. In addition, 3 studies contributed almost 3 quarters of the data for evaluation of the primary endpoint, 30-day postoperative mortality. Randomized control trials to compare the short- and long-term outcomes from CE + CABG versus isolated CABG are still lacking. Thus, the limitations associated with observational studies, such as difficulties to draw causal inferences and the potential biased effects of confounding factors on study endpoints, cannot be avoided. In addition, most of the included observational studies are small-scale studies, and only 4 studies contain relatively large sample size (more than 1000 participants). Meta-analysis including high-quality prospective randomized control studies in addition to observational studies is required to further confirm the findings of this present study. Furthermore, because of limited data availability, only a few short-term postoperative complications including 30-day postoperative MI, AF, VT, and renal dysfunction were examined in this study. The effects of CE + CABG on other postoperative complications need to be investigated in future studies. Long-term outcomes in addition to survival, such as incidence of long-term MI and angina, should be also investigated in future. Last, in the subgroup analysis of patients with diffuse disease in the LAD, 3 of the 5 included articles were from the same group (Dr Takanashi's group). Thus, future large-scale studies are required to verify the conclusion.

CONCLUSIONS

In this study, we found that CE + CABG was associated with significantly increased 30-day postoperative mortality and higher incidences of short-term postoperative complications compared with isolated CABG, particularly in the subgroup patients with high-risk profile and patients with diffuse disease in the LAD. Although the comparable long-term survival in CE + CABG and isolated CABG groups may indicate beneficial effects of complete revascularization by CE + CABG on patients requiring CE, further investigations are required to verify this conclusion. Thus, to achieve satisfactory efficacy, surgeons should carefully weigh possible benefits and adverse effects of CE and prepare the surgical strategy adequately.

TABLE 1 (Continued)

Characteristics of the Included Studies

TABLE 2

Long-Term Results

TABLE 3

Meta-Analysis on Short-Term Postoperative Complications Including All the Eligible Reports

TABLE 4

Meta-Analysis on Short-Term Complications Including the Recent Reports (Published Later Than 2000)